Bitzu

AI-901 ยท Common AI workloads

What does each core text analysis technique return?

Keyword (key phrase) extraction, entity detection, sentiment analysis and summarization: what each returns and which fits a scenario.

Key points

  • Key phrase extraction: a list of the main concepts.
  • Entity detection (named entity recognition): spans of text labeled with a type such as person, place, organization or date.
  • Sentiment analysis: how positive or negative the text is, with confidence scores (opinion mining links opinions to specific aspects).
  • Summarization: a shorter version, either extractive (selected sentences) or abstractive (newly written sentences).

Source: https://learn.microsoft.com/en-us/azure/ai-services/language-service/overview

Practice questions

Single choice

Which technique returns the main concepts of a document as a simple list?

  1. Sentiment analysis
  2. Language detection
  3. Key phrase extraction
  4. Entity detection
Show answer

Answer: C. Key phrase extraction

Key phrase extraction returns the main concepts as a list. Entity detection returns typed items, sentiment returns opinion scores, and language detection returns the language of the text.

Select all that apply

Which statements about text summarization are correct?

  1. Extractive summarization selects key sentences from the source
  2. Abstractive summarization writes new, concise sentences
  3. Summarization can also be applied to conversations and call transcripts
  4. Summarization returns a list of typed entities
  5. Summarization needs audio input
Show answer

Answer: A. Extractive summarization selects key sentences from the source, B. Abstractive summarization writes new, concise sentences, C. Summarization can also be applied to conversations and call transcripts

Extractive summaries select existing sentences; abstractive ones write new sentences. Summarization also covers conversations such as chats and call transcripts. It does not return typed entity lists.

True or false

True or false: sentiment analysis tells you which topics a review talks about.

Show answer

Answer: False

False. Sentiment analysis gauges how positive or negative the text is. Topics are surfaced by key phrase extraction.

Practice 8 questions on this topic in Bitzu

Short lessons, instant explanations and spaced review of what you get wrong. Free on iOS and Android.

Download on the App StoreGet it on Google Play

More AI-901 topics in this module